A widely distributed large scale computer architecture is a highly complex system that typically includes hundreds if not thousands of physical and/or virtual machines. Operational issues inevitably arise due in part to the large number and distribution of computing resources throughout the distributed computer architecture. Operational problems may be caused by significant hardware memory and disk space usage, disk errors, network hardware issues, insufficient or misconfigured hardware and/or software components, malfunctioning hardware and/or software components, etc.
Attempts have been made to implement focused operational procedures and/or parameters in order to maintain a well-functioning and responsive system where back and forth communications occur smoothly. For example, each worker machine of an array of worker node machines may utilize a local throttling technique in order to regulate incoming and/or outgoing network traffic and/or limit bandwidth congestion. A machine may employ embedded throttling logic that determines which applications or processes are allowed to use CPU and/or memory resources including rules that determine which application or process should execute before other applications or processes.
Local throttling attempts may give rise to bottlenecks that result in delays in sending, receiving, and/or processing electronic data. For example, a calling application's internal throttling logic may be employed when attempting to avoid releasing a flood of electronic data which may result in certain types of electronic data inadvertently being held back. The application is not aware of current processing issues throughout the system and thus has to artificially and crudely limit throughput. It is not uncommon for thousands of electronic transactions to be released to a clearinghouse computer system which, even using an application's internal throttling logic, may result in communication throughput and bottleneck issues. Local throttling may result in communication bottlenecks that may unduly limit performance of an associated machine and ultimately affect performance of the widely distributed large scale computer architecture. Unfortunately, the net impact is that bottlenecks result in delayed transaction processing, delayed queries, and delayed data access requests.